Bonds were roughly unchanged overnight but soon began to improve at the CME and NYSE opening bells.  Weak economic data at 10am didn't hurt the rally, but it didn't help as much as more important data would have.  In the afternoon hours, speeches from the Fed's Bullard and Powell put some upward pressure on rates as they reminded markets that it's not all gloom and doom when it comes to considering reasons that the Fed may (or may not) cut rates in July. 

Ultimately, however, the presence of the risks combined with Powell's reminder that it made more sense to stay well in front of any potential downturn was enough to preserve most of the gains from the morning hours. 

All of the above is well and good, but the "gains from the morning hours" is a phrase best applied to the longer end of the Treasury curve.  MBS didn't get nearly as much love by comparison.  This is getting to be a familiar refrain over the past few months, but it's not an abnormal or surprising one. 

MBS prices and especially MBS spreads vs Treasuries are significantly impacted by significant volatility.  Given that we're in the midst of a large-scale repricing of interest rate realities (and a large-scale re-discovery of the effects on refi/purchase demand)--AND that there are incredibly significant events coming up in the next few days--makes the underperformance logical enough.  At some point, MBS will have been shunned enough to account for all of the variables and the ratchet will turn the other way.